Our long term goal is to understand the role in breast cancer of retinoic acid receptor (RAR) genes and genes that function upstream of RAR to regulate vitamin A storage and retinoic acid biosynthesis, in particular the cellular retinol-binding protein 1 gene (CRBP1). The rationale for this goal is the growth suppressive and differentiation-inducive activities of RARs and the downregulation in breast cancer of RAR isoforms and CRBP1, Using transgenic mouse models that are relevant to human breast cancer, we found that RARs modulate tumor incidence in an oncogene-specific manner. Our data predict that retinoid intervention will be beneficial or counterproductive depending on the oncogenic pathway driving transformation. In Aim 1, we will test our specific hypothesis that treatment of mammary tumor-prone MMTV- Neu females with retinoid (Am580/Am80) will accelerate tumor development whereas the same treatment of mammary tumor-prone MMTV-Wnt1 and MMTV-Myc females will delay tumor development. In related studies, we found that CRBP1 promotes vitamin A storage in breast tumor epithelial cells and positively regulates RAR activity;in turn, RAR inhibits the PI3K/Akt survival pathway promoting tumor cell differentiation and growth inhibition. Therefore, CRBP1 links cellular uptake and storage of a dietary nutrient, vitamin A, to cell differentiation and growth suppression. We found that, just as in human breast cancer, CRBP1 is downregulated in the MMTV-Myc mouse model of breast cancer. In Aim 2, we will study MMTV- Myc transgenic mice to test whether loss of CRBP1 expression is a causal event that precedes tumor formation. In Aim 3, we will generate MMTV-CRBP1 transgenic mice and cross these to MMTV-Myc mice to test whether ectopic CRBP1 expression prevents tumor formation by Myc. New preventive drugs are needed for women with early estrogen receptor negative breast cancer. We propose that selective retinoids can be effective but only when the disease is associated with specific cancer causing genes (it may otherwise prove deleterious). If our concept is validated, it will guide the design of retinoid clinical trials. We will also critically address the relevance of a vitamin A-binding protein that is lost in breast cancer. We believe that this protein defines whether breast cells can use vitamin A and that its loss contributes to disease causation or progression.